Estimating net ecosystem exchange in a patterned ecosystem: Example from blanket bog

Net ecosystem exchange (NEE) was measured in a patterned peatland with eddy covariance (EC) and chamber methods during a 12-month period. The peatland surface was composed of microforms characterized by a difference in water level and vegetation composition. The distribution of microforms varied spatially within the peatland. To achieve correspondent half-hourly EC and chamber NEE estimates, we modelled microform level chamber fluxes, estimated them at each instance of weather recordings and integrated them over the year. We then scaled the fluxes up to the EC footprint. On a half-hourly time scale the correlation coefficient (R) of the NEE between the methods was 0.80 and the slope of regression 0.91. Measurements made in summer and during daylight were more highly correlated than measurements made in winter and during darkness. When integrated on a monthly time scale the methods agreed better, with R of 0.98 and slope of regression 1.00. The annual NEE for the EC and chamber methods were 206 and 242 g(CO2) m−2, respectively. The study confirmed that the surface pattern of the EC footprint in the blanket bog was sufficiently homogeneous, that the changing wind direction did not influence the half-hourly NEE. However, the chamber estimates found that the annual NEE of the driest area within the footprint was 130% larger than that in the wettest area, indicating that large spatial variation can be found in NEE.